HickoryDnsResolver use a shared instance by default to limit fd use (#5523)

This commit is contained in:
Jack Wampler
2025-02-27 09:05:10 -07:00
committed by GitHub
parent 856dbfe1ac
commit f13ce6bf2d
3 changed files with 95 additions and 17 deletions
+89 -11
View File
@@ -13,15 +13,18 @@
use crate::ClientBuilder;
use std::{net::SocketAddr, sync::Arc};
use std::{
net::SocketAddr,
sync::{Arc, LazyLock},
};
use hickory_resolver::lookup_ip::LookupIp;
use hickory_resolver::{
config::{LookupIpStrategy, NameServerConfigGroup, ResolverConfig, ResolverOpts},
error::ResolveError,
lookup_ip::LookupIpIntoIter,
TokioAsyncResolver,
};
use hickory_resolver::{error::ResolveErrorKind, lookup_ip::LookupIp};
use once_cell::sync::OnceCell;
use reqwest::dns::{Addrs, Name, Resolve, Resolving};
use tracing::warn;
@@ -38,6 +41,14 @@ struct SocketAddrs {
iter: LookupIpIntoIter,
}
// n.b. static items do not call [`Drop`] on program termination, so this won't be deallocated.
// this is fine, as the OS can deallocate the terminated program faster than we can free memory
// but tools like valgrind might report "memory leaks" as it isn't obvious this is intentional.
static SHARED_RESOLVER: LazyLock<HickoryDnsResolver> = LazyLock::new(|| {
tracing::debug!("Initializing shared DNS resolver");
HickoryDnsResolver::default()
});
#[derive(Debug, thiserror::Error)]
#[error("hickory-dns resolver error: {hickory_error}")]
/// Error occurring while resolving a hostname into an IP address.
@@ -47,29 +58,62 @@ pub struct HickoryDnsError {
}
/// Wrapper around an `AsyncResolver`, which implements the `Resolve` trait.
///
/// Typical use involves instantiating using the `Default` implementation and then resolving using
/// methods or trait implementations.
///
/// The default initialization uses a shared underlying `AsyncResolver`. If a thread local resolver
/// is required use `thread_resolver()` to build a resolver with an independently instantiated
/// internal `AsyncResolver`.
#[derive(Debug, Default, Clone)]
pub struct HickoryDnsResolver {
/// Since we might not have been called in the context of a
/// Tokio Runtime in initialization, so we must delay the actual
/// construction of the resolver.
// Since we might not have been called in the context of a
// Tokio Runtime in initialization, so we must delay the actual
// construction of the resolver.
state: Arc<OnceCell<TokioAsyncResolver>>,
fallback: Arc<OnceCell<TokioAsyncResolver>>,
dont_use_shared: bool,
}
impl Resolve for HickoryDnsResolver {
fn resolve(&self, name: Name) -> Resolving {
let resolver = self.state.clone();
let fallback = self.fallback.clone();
let independent = self.dont_use_shared;
Box::pin(async move {
let resolver = resolver.get_or_try_init(new_resolver)?;
let resolver = resolver.get_or_try_init(|| {
// using a closure here is slightly gross, but this makes sure that if the
// lazy-init returns an error it can be handled by the client
if independent {
new_resolver()
} else {
Ok(SHARED_RESOLVER.state.get_or_try_init(new_resolver)?.clone())
}
})?;
// try the primary DNS resolver that we set up (DoH or DoT or whatever)
let lookup = match resolver.lookup_ip(name.as_str()).await {
Ok(res) => res,
Err(e) => {
// on failure use the fall back system configured DNS resolver
warn!("primary DNS failed w/ error {e}: using system fallback");
let resolver = fallback.get_or_try_init(new_resolver_system)?;
match e.kind() {
ResolveErrorKind::NoRecordsFound { .. } => {}
_ => {
warn!("primary DNS failed w/ error {e}: using system fallback");
}
}
let resolver = fallback.get_or_try_init(|| {
// using a closure here is slightly gross, but this makes sure that if the
// lazy-init returns an error it can be handled by the client
if independent {
new_resolver_system()
} else {
Ok(SHARED_RESOLVER
.fallback
.get_or_try_init(new_resolver_system)?
.clone())
}
})?;
resolver.lookup_ip(name.as_str()).await?
}
};
@@ -93,21 +137,55 @@ impl Iterator for SocketAddrs {
impl HickoryDnsResolver {
/// Attempt to resolve a domain name to a set of ['IpAddr']s
pub async fn resolve_str(&self, name: &str) -> Result<LookupIp, HickoryDnsError> {
let resolver = self.state.get_or_try_init(new_resolver)?;
let resolver = self.state.get_or_try_init(|| self.new_resolver())?;
// try the primary DNS resolver that we set up (DoH or DoT or whatever)
let lookup = match resolver.lookup_ip(name).await {
Ok(res) => res,
Err(e) => {
// on failure use the fall back system configured DNS resolver
warn!("primary DNS failed w/ error {e}: using system fallback");
let resolver = self.fallback.get_or_try_init(new_resolver_system)?;
match e.kind() {
ResolveErrorKind::NoRecordsFound { .. } => {}
_ => {
warn!("primary DNS failed w/ error {e}: using system fallback");
}
}
let resolver = self
.fallback
.get_or_try_init(|| self.new_resolver_system())?;
resolver.lookup_ip(name).await?
}
};
Ok(lookup)
}
/// Create a (lazy-initialized) resolver that is not shared across threads.
pub fn thread_resolver() -> Self {
Self {
dont_use_shared: true,
..Default::default()
}
}
fn new_resolver(&self) -> Result<TokioAsyncResolver, HickoryDnsError> {
if self.dont_use_shared {
new_resolver()
} else {
Ok(SHARED_RESOLVER.state.get_or_try_init(new_resolver)?.clone())
}
}
fn new_resolver_system(&self) -> Result<TokioAsyncResolver, HickoryDnsError> {
if self.dont_use_shared {
new_resolver_system()
} else {
Ok(SHARED_RESOLVER
.fallback
.get_or_try_init(new_resolver_system)?
.clone())
}
}
}
/// Create a new resolver with a custom DoT based configuration. The options are overridden to look